• Generated millions in revenue by leading the design, planning, and execution of Identity Federation support for four Dialogflow consoles, driving cross-functional teams to ensure timely and accurate delivery.

• Ensured 99.9% service uptime by guiding cross-functional alignments, defining SLOs, and launch of a token-based LLM rate limiting system for Dialogflow using C++, Java, Kotlin, and Guice.

• Led the development of the Travel AI MVP, a Generative AI travel planning tool. Defined team priorities and architected core infrastructure, LLM agent, integrations with YouTube and Places APIs, and CI/CD pipelines for version control and deployment using GCP, Python, Terraform, Vertex AI, Dialogflow, and FastAPI.

• Drove the GA release of AML AI, a GCP product generating millions in revenue; built key features, identified and resolved GA-blocking issues, and boosted profits via billing enhancements using Go and BigQuery.

• Elevated customer experience by designing and integrating compliance monitoring systems, and enhanced product's reliability through comprehensive testing strategy using Go, Python, Bazel, and Protobuf.

Spearheaded the development, testing, and launch of new Contract Analysis models, collaborating with ML scientists and integrating the models into the production infrastructure using Airflow and AWS ECS.

• Increased operational efficiency by 50% through addressing disparities between new and old stacks, deprecating the old stack, and simplifying infra, while releasing updates to customers using Python and AWS.

• Led the design and development of a distributed data ingestion and processing system on AWS, enabling $20M in potential revenue by providing access to business-level data for analysis and innovation; automated the CI/CD infrastructure using Terraform, Jenkins, Docker, and Airflow.

• Proposed a millimeter wave-based road condition sensing machine learning model resulted in 10× lower cost and 15% accuracy improvement; collected and cleaned a dataset of 80k+ samples and trained the model using teacher-student network using Python, C++, and Scikit-Learn.

• Implemented a data transmission system with 3× more bandwidth and 70× lower cost using C++ & Python.